The objectives of the proposed research are: to test the validity of current methods of reconstructing protein phylogenies; to test whether specific hypotheses about patterns of protein variation that have been derived from reconstructions (in particular, the molecular clock hypothesis, the apparent evolutionary increase in informational entropy, and non-random patterns of nucleotide substitution) may be attributed to artifacts of reconstruction; to evaluate measures of evolutionary distance; to develop more accurate and less biased methods of reconstruction and distance estimation, and to derive newly discernible patterns of evolutionary change. The principal method used will be computer simulation of protein phylogeny and evolution. Input variables will include various non- uniform distributions of the probability of change at individual amino acid sites; topology and leg lengths; transition probabilities for each pair of amino acids, as derived from reconstructions of real data; and nucleotide frequencies as derived from real data. Terminal sequences of simulated phylogenies will be used to build reconstructed phylogenies with hypothetical ancestral sequences, leg lengths, transition probabilities, etc. according to currently accepted reconstruction algorithms. Various distance measures will be calculated between all pairs of terminal sequences. The reconstructed phylogenies and calculated distance measures will be compared with the input (simulated) phylogenies and distances, with particular attention given to whether patterns previously reported from the reconstructed phylogenies of real protein families also appear in the reconstructed phylogenies of simulated protein families, but not in the input (simulated) phylogenies themselves. Parameters of simulation will be progressively adjusted to most closely fit the variation seen in real protein families, which will be subjected to renewed analysis. Hopefully not only will spurious and artifactual patterns be discredited and valid hypotheses strengthened, but valid new patterns will be discerned.